ARTICLE IN PRESS

Medical Dosimetry ■■ (2018) ■■–■■

Medical Dosimetry
j o u r n a l h o m e p a g e : w w w. m e d d o s . o r g

Dosimetry Contribution:

Offline adaptive radiation therapy in the treatment of

prostate cancer: a case study
Evgenia Nigay, M.S., R.T.(T), Heath Bonsall, M.S., C.M.D., R.T.(T),
Beverly Meyer, M.S., R.T.(R)(T), Ashley Hunzeker, M.S., C.M.D., and
Nishele Lenards, M.S., C.M.D., R.T.(R)(T), F.A.A.M.D.
Medical Dosimetry Program at the University of Wisconsin, La Crosse, WI

A R T I C L E I N F O A B S T R A C T

Article history: The purpose of this case study is to develop a method to account for the difference in the
Received 21 October 2017 daily volumes in the bladder, rectum, and targets in prostate radiotherapy and to compare
Received in revised form 18 the predicted dose to the actual dose to these organs. Five patients, both prospectively and
December 2017
retrospectively, were selected from 2 different cancer centers, with a biopsy-confirmed di-
Accepted 18 December 2017
agnosis of prostate cancer. The patients’ planning target volume (PTV) and organs at risk
(OAR) were contoured on the computed tomography (CT) dataset using either Eclipse or
Keywords:
Monaco treatment planning systems (TPSs). Cone-beam computed tomography (CBCT) scans
Adaptive radiation therapy
were collected before each daily treatment and exported to MIM software for analysis. The
prostate cancer
automatically generated reports evaluated the organ volume changes, the actual dose re-
MIM software
ceived during a single fraction, and the projected dose to each organ at the completion of
the treatment course via comparative cumulative dose-volume histograms (DVHs). Volume
changes in the bladder and rectum can cause notable variations in the prescribed dose vs
the actual dose received. MIM software was proven to have utility prospectively by tabu-
lating daily dose and projecting final doses, potentially aiding physicians in decisions about
the boost plans, thus making offline adaptive radiation therapy (ART) clinically manageable.
© 2018 American Association of Medical Dosimetrists.

Introduction
Among men in the United States, prostate cancer is the
second most common cancer and one of the most common
causes of death. 1 Radiation therapy, both external and
brachytherapy, for prostate cancer is a highly utilized form
of treatment in cancer centers throughout the United States.
With external radiotherapy, doses required for tumor
Reprint requests to Evgenia Nigay, M.S., R.T.(T), Medical Dosimetry
Program at the University of Wisconsin, La Crosse, WI.
E-mail: evgenia.nigay@gmail.com
response can reach 70 Gy. Due to such high doses, delivery
of a sufficient dose to the prostate or prostate bed while lim-
iting dose to adjacent radiosensitive structures, such as the
bladder and rectum, proves to be difficult.2 Organs at risk
(OAR) are highly mobile, morphing shapes that can make
dose delivery highly speculative, creating questionable dose
to the OAR. Even as sophisticated as the plans have become
for radiation therapy, they cannot account for daily changes
in patient anatomy.3
Adaptive radiation therapy (ART) has been around for
decades.3 However, technical complications can make ART
difficult to come to fruition, namely, the timely efforts in re-
contouring and re-planning each patient who is receiving
radiation therapy treatment. 3 Cone beam computed

https://doi.org/10.1016/j.meddos.2017.12.005
0958-3947/Copyright © 2018 American Association of Medical Dosimetrists
 ARTICLE IN PRESS
2 H. Bonsall et al. / Medical Dosimetry ■■ (2018) ■■–■■
tomography (CBCT) is considered to be the standard check
utilized for online ART before daily radiation treatments. As
studies have shown, the use of CBCT can decrease the acute
and chronic side effects to the genitourinary and gastroin-
testinal systems.4
ART can play an important role in decreasing unwanted
dose to critical structures and normal tissues. Prostate bed
and intact prostate radiation treatments are heavily reliant
upon proper and consistent bladder filling as well as rectal
emptying. When these 2 criteria are not adequately repro-
duced according to the computed tomography (CT) planning
scan, the dose delivered to the target can be considerably
altered.2 Basing the plan on 1 CT scan, daily changes in patient
positioning by even a few millimeters can result in devia-
tion from the intended dose.4 The International Commission
on Radiological Units and Measurements recommends
regular reports be completed on the received dose to the
planning OAR volume, which is now available through the
use of daily CBCT.5
Daily changes in positions of OAR and their respective
volumes can alter dose considerably, warranting investiga-
tion. Retrospectively analyzing the dose to the target can
reveal a high fluctuation in minimum dose to the prostate.6
To track the daily doses to the target and OAR, MIM soft-
ware (Cleveland, OH) has been utilized.7 MIM software is a
suite of tools that allows for efficient contouring, rigid and
deformable registration, and plan evaluation. It has been uti-
lized in an accurate and automated ART workflow in
conjunction with methods that use deformable CBCT con-
tours. In this case study, the adaptive radiotherapy assistant
automated the process of fusing the daily CBCT to the plan-
ning CT, deforming the dose and the contours, and generating
reports of the findings. The goal of this research study was
to develop a method to account for the actual difference in
the volumes from day to day and to compare the predicted
dose to the actual dose delivered to the bladder, rectum, and
targets. The workflows were designed to edit the trans-
ferred contours and evaluate them as part of the process.
Case Description
Patient selection
Patients for this study were selected in a retrospective as
well as prospective manner. Each patient was diagnosed with
prostate adenocarcinoma confirmed via biopsy. Selected pa-
tients were to have the external radiotherapy treatments to
the intact prostate with or without nodal involvement or to
the prostate bed.
For the simulation, patients were placed in the supine po-
sition with a pillow under the head and arms high on their
chest, holding a ring. The legs were immobilized in a Vac-
Lok (CIVCO Radiotherapy, Coralville, IA). Each patient received
a set of permanent marks in the pelvic region, anteriorly and
on either side, to provide a reference point for treatment
planning. Radiopaque markers were placed on these marks
to visualize them during the scan. Patients were scanned
head first, using 2-mm slices on either Siemens SOMATOM
Emotion 16 or Siemens SOMATOM Definition AS (Simens,
Erlangen, Germany) CT scanner.
As per the clinic’s protocol, the patients were instructed
to empty their rectum and fill their bladder before the sim-
ulation and each day before treatment. If the rectum was
too full, patients were asked to have a bowel movement
before repeating the scan. Patients were also instructed to
have a low-residue diet for the duration of the treatment to
help prevent or reduce diarrhea.
Target delineation
Once the simulation scan was complete, the target de-
lineation and treatment planning were performed using
either the Elekta Monaco 5.11.01 TPS (Elekta, Stockholm,
Sweden) or Varian Eclipse 13.7 TPS (Varian, Palo Alto, CA).
The clinical target volume (CTV) was contoured by the phy-
sician on the CT dataset. The planning target volume (PTV)
was created by expanding the CTV by 3 mm posteriorly and
5 mm in all other directions for the intact prostate or 8 mm
in all directions for the prostate fossa. Once the CT dataset
was received by the medical dosimetrist, OAR were con-
toured. The OAR of interest for this study included the rectum
and bladder volumes.
Before each treatment, the patients were imaged using
CBCT. The scans were matched to the initial planning CT scan
to ensure the precise target location as well as consistency
in bladder filling and rectal emptying during each treat-
ment. Each site aligned the CBCTs based on the priorities
provided by the respective attending physician. Generally,
patients were aligned to implanted fiducials if they were
present. If nodal volumes or prostate fossa were being
treated, the local bony anatomy was used. The CBCT scans
along with any shifts made were sent to the MIM software
for analysis.
Treatment planning
Specified by the physician, the treatment goals included
uniform coverage of the PTV while keeping the dose to the
bladder and rectum as low as possible. The prescription dose
varied depending on whether the target was the intact pros-
tate or the prostate fossa. The treatment plans were created
using either the volumetric arc therapy or static intensity-
modulated radiation therapy. Patient no. 1 was treated to
the intact prostate using a 2-full arc volumetric arc therapy.
Patients 2 through 5 were treated using static intensity-
modulated radiation therapy with a 9-beam arrangement.
 ARTICLE IN PRESS
H. Bonsall et al. / Medical Dosimetry ■■ (2018) ■■–■■ 3

Table 1
Prescription details for the initial plan for patient nos. 1-5

Site Patient no. 1 Patient no. 2 Patient no. 3 Patient no. 4 Patient no. 5
Prostate Prostate fossa Prostate Prostate Prostate
Prescription dose 54 Gy/30 fx 50.4 Gy/28 fx 45 Gy/25 fx 50 Gy/20 fx 45 Gy/25 fx
Planning technique VMAT, 2 full arcs IMRT, 9 beams IMRT, 9 beams IMRT, 9 beams IMRT, 9 beams

The prescription for each patient is summarized in Table 1.
For the purpose of this study, only the initial plans without
the boost were used.
Desired dose to critical structures was not to exceed the
guidelines set forth by the Quantitative Analyses of Normal
Tissue Effects in the Clinic (QUANTEC) and the Radiation
Therapy Oncology Group (RTOG). Constraints by the RTOG
were utilized for both prostate and prostate fossa plans; ad-
ditionally, QUANTEC constraints were used for the prostate
fossa plans. As per QUANTEC constraints, volume of rectum
receiving 50 Gy (V50) cannot exceed 50% of the total organ
volume (V50 ≤ 50%), volume receiving 60 Gy cannot exceed
35% (V60 ≤ 35%), and volume receiving 65 Gy cannot exceed
25% (V65 ≤ 25%). For the bladder, volume receiving 65 Gy
cannot exceed 50% of the organ volume (V65 ≤ 50%) and
volume receiving 70 Gy cannot exceed 35% (V70 ≤ 35%). Fol-
lowing the RTOG constraints, volume of the rectum receiving
40 Gy cannot exceed 55% (V40 ≤ 55%) and volume of the
bladder receiving 40 Gy cannot exceed 70% (V40 ≤ 70%). The
summary of the dose-volume constraints can be found in
Table 2.
Table 2
Constraints from QUANTEC and RTOG for OAR
QUANTEC
Bladder V65 ≤ 50%
V70 ≤ 35%
Rectum V50 ≤ 50%
V60 ≤ 35%
V65 ≤ 25%
For each patient, the daily CBCT was completed and
exported into MIM software, where the assistant looked
for the daily CBCTs as they were exported. Once it recog-
nized a new CBCT, MIM automatically ran through a
workflow that performed a deformable registration with
dose and contour transfer. Another workflow was then used
by the reviewing medical dosimetrist to go through each
of the deformed contours and make edits where neces-
sary. The software then adjusted and generated the report
of the findings. These reports provided information on the
organ volume changes, planning dose to each organ com-
pared to the actual dose received during a single fraction,
and the projected dose to each organ at the completion of
the treatment course.
Plan analysis and evaluation
Patient no. 1 was analyzed retrospectively for all 30 treat-
ment fractions. The comparative cumulative dose-volume
histograms (DVHs) showed no significant variations in the
actual delivered dose when compared to the planning dose
(Fig. 1). The PTV coverage was slightly less than antici-
pated, with 95.85% of the volume being covered by 95% of
the prescription dose vs 98.07% of the volume covered by
RTOG
95% of the prescription dose during the initial planning.
V40 ≤ 70%
Overall, the actual dose to the bladder was slightly lower than
V40 ≤ 55% the planned dose, and the actual dose to the rectum was
slightly higher than the planned dose but without signifi-
cant variation. Doses to the CTV were also comparable.

Fig. 1. Patient no. 1—cumulative DVH showing planned dose vs total delivered dose; 30 fractions (fx). (Color version of figure is available online.)
 ARTICLE IN PRESS
4 H. Bonsall et al. / Medical Dosimetry ■■ (2018) ■■–■■

Fig. 2. Patient no. 2—cumulative DVH showing planned dose vs total delivered dose; 28 fx. (Color version of figure is available online.)

Patient no. 2 was analyzed while on treatment. The com-
parative cumulative DVH showed actual dose delivered to
the CTV to be comparable to the planning dose (Fig. 2). The
PTV received lower dose than planned, with 95% of the dose
received by 95.53% of the volume vs the planned 98.29% of
the volume. The dose received by the rectum was consis-
tent during daily treatments but was overall higher than the
initially planned dose. The dose to the bladder was compa-
rable and slightly lower than the planned dose.
The comparative cumulative DVH for patient no. 3, who
was analyzed while on treatment, showed a comparable cov-
erage of the CTV and the PTV (Fig. 3). The actual dose to 15%
of the rectum was 36.75 Gy, an increase from the planned
dose of 32.91 Gy. The rectal volume overall also received a
higher dose than planned. The dose received by the bladder
was very comparable.
Patient no. 4 was analyzed while on treatment using a
hypo-fractionated prescription and showed no deviation in
CTV coverage on the comparative cumulative DVH (Fig. 4).
The PTV coverage was reduced from 96.66% of the volume
covered by the prescription dose to 91.25%. The bladder
received the dose that was similar to the planned dose. The
dose to the rectal volume was higher than planned, with 15%
of the volume receiving 43.36 Gy vs 37.47%.
Patient no. 5 was analyzed while on treatment. The CTV
and the PTV coverage were comparable between the planned
dose and actual received dose (Fig. 5). The cumulative dose
to the bladder volume also did not deviate from the initial-
ly planned dose. Rectal volume, however, received a higher-
than-anticipated dose. The dose to 15% of the rectal volume
was 43.29 Gy, an increase from 40.55 Gy.
The reports were generated for each daily fraction for all
the patients. The DVH created showed the variations between
planning doses and actual single fraction doses. Daily varia-
tions of dose received by the OAR were recorded. Overall,
the dose received by the OAR on a daily basis fluctuated only
slightly, with more notable changes recorded in the dose re-
ceived by the rectal volume. In all 5 patients, the greatest
variation in dose for the 15% of the rectal volume was 15.72%
higher than the planned dose. The dose received by 15% of
the bladder volume was 3% lower than the planned dose.
Daily variations in OAR volumes were also recorded with

Fig. 3. Patient no. 3—cumulative DVH showing planned dose vs total delivered dose; 25 fx. (Color version of figure is available online.)
 ARTICLE IN PRESS
H. Bonsall et al. / Medical Dosimetry ■■ (2018) ■■–■■ 5

Fig. 4. Patient no. 4—cumulative DVH showing planned dose vs total delivered dose; 20 fx. (Color version of figure is available online.)

significant fluctuations from fraction to fraction. Fig. 6 shows
the variations in bladder volume between the initial plan-
ning CT and each daily fraction for all 5 patients. Fig. 7 shows
the daily variations in rectal volume for all patients.
Conclusion
For each patient, there were notable variations when com-
paring planning dose with the actual dose delivered to the
target and OAR, especially the rectum daily dose. The daily
changes in turn add up to a cumulative dose that differs from
the initial planning dose. These findings prove that there is
utility in using the MIM software to track the changes. When
completed while the patients are on treatment, the entire
process from uploading the daily CBCT scan into the soft-
ware to editing the contours and generating the reports takes
between 5 and 10 minutes. Analyzing all the CBCTs retro-
spectively was proven to be too time-consuming. This renders
the software most useful for the patients who are cur-
rently on treatment.
The data collected were for research purposes and for eval-
uation of the efficiency of the offline ART process, and were
therefore not reviewed by physicians. For further research,
physicians can be involved to determine whether the results
are clinically significant, as dose tracking during the initial
portion of the plan will provide useful information when de-
signing the boost plan. Because the actual delivered dose will
be available from the initial plan, it will provide the planner
with a better idea of how much of the dose is still allow-
able to the OAR. The target coverage can also be adjusted
for the boost plan if it was not adequate during the initial
plan. In addition, knowing the actual doses received might
aid physicians in deciding whether further dose escalation
is feasible or if the OAR would be overdosed. The next step
in utilizing the software could be to re-optimize the plan
before the daily treatment if the variations are significant
enough.
Because the software monitors daily doses, it takes each
fraction into account and not only compares the daily varia-
tions, but also creates a projected final dose. This function
of MIM software makes analyzing the dose delivered on a
daily basis very useful. It makes possible to detect changes
that are so significant that they might prompt another plan-
ning scan and a new treatment plan. Despite some of the

Fig. 5. Patient no. 5—cumulative DVH showing planned dose vs total delivered dose; 25 fx. (Color version of figure is available online.)
 ARTICLE IN PRESS
6 H. Bonsall et al. / Medical Dosimetry ■■ (2018) ■■–■■

Fig. 6. Daily variations in bladder volume for patient nos. 1-5 (fx. 0 = initial planning volume). (Color version of figure is available online.)

software limitations, such as the need to manually edit de- manageable. Based on the study findings, the changes dis-
formable contours, the workflows designed by the MIM played during daily treatments as well as the actual
software engineers make the process of offline adaptive cumulative dose variations justify the extra workload that
therapy as automated as possible, making it clinically goes into analyzing the daily CBCTs with MIM software.

Fig. 7. Daily variations in rectal volume for patient nos. 1-5 (fx. 0 = initial planning volume). (Color version of figure is available online.)
 ARTICLE IN PRESS
H. Bonsall et al. / Medical Dosimetry ■■ (2018) ■■–■■
References
1. Prostate Cancer Statistics. Centers for Disease Control and Prevention
Web site. Updated May 23, 2017. Available at: https://www.cdc.gov/
cancer/prostate/statistics. Accessed June 21, 2017.
2. Huang, T.C.; Chou, K.T.; Yang, S.N.; et al. Fractionated changes in
prostate cancer radiotherapy using cone-beam computed tomography.
Med. Dosim. 40(3):222–5; 2015. http://dx.doi.org/10.1016/j.meddos
.2014.12.003.
3. Lim-Reinders, S.; Keller, B.M.; Al-Ward, S.; et al. Online adaptive
radiation therapy. Int. J. Radiat. Oncol. Biol. Phys. 99(4):994–1003; 2017.
http://dx.doi.org/10.1016/j.ijrobp.2017.04.023.
4. Tonlaar, N.Y.; Marina, O.; Brabbins, D.S.; et al. Use of weekly cone beam
CT for adaptive radiation therapy decreases toxicity for prostate cancer
7
patients treated with IMRT. Int. J. Radiat. Oncol. Biol. Phys. 87(2,
suppl.):S176; 2013. http://dx.doi.org/10.1016/j.ijrobp.2013.06.454.
5. Prabhakar, R.; Oates, R.; Jones, D.; et al. A study on planning organ
at risk volume for the rectum using cone beam computed tomography
in the treatment of prostate cancer. Med. Dosim. 39(1):38–43; 2014.
http://dx.doi.org/10.1016/j.meddos.2013.09.003.
6. Noel, C.E.; Santanam, L.; Olsen, J.R.; et al. An automated method for
adaptive radiation therapy for prostate cancer patients using
continuous fiducial-based tracking. Phys. Med. Biol. 55(1):65–82; 2010.
http://dx.doi.org/10.1088/0031-9155/55/1/005.
7. Pirozzi, S.; Piper, J.; Nelson, A.; et al. Evaluation of deformable prostate
cone beam computed tomography (CBCT) contouring methods for
adaptive radiation therapy. Int. J. Radiat. Oncol. Biol. Phys. 87(2,
suppl.):S719; 2013. http://dx.doi.org/10.1016/j.ijrobp.2013.06.1904.